TheBloke commited on
Commit
8314976
1 Parent(s): b5311b9

Upload new k-quant GGML quantised models.

Browse files
Files changed (1) hide show
  1. README.md +78 -25
README.md CHANGED
@@ -1,9 +1,8 @@
1
  ---
 
2
  license: other
3
- datasets:
4
- - c-s-ale/alpaca-gpt4-data
5
- pipeline_tag: text2text-generation
6
  ---
 
7
  <!-- header start -->
8
  <div style="width: 100%;">
9
  <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
@@ -18,37 +17,86 @@ pipeline_tag: text2text-generation
18
  </div>
19
  <!-- header end -->
20
 
21
- ## GPT4-Alpaca-LoRA_MLP-65B GGML
22
 
23
- These files are the result of merging the [LoRA weights of chtan's gpt4-alpaca-lora_mlp-65B](https://huggingface.co/chtan/gpt4-alpaca-lora_mlp-65b) with the original Llama 65B model.
24
 
25
- This repo contains 4bit and 5bit quantised GGML files for CPU inference using [llama.cpp](https://github.com/ggerganov/llama.cpp).
 
 
 
 
 
26
 
27
  ## Repositories available
28
 
29
- * [4bit GPTQ models for GPU inference](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-GPTQ)
30
- * [4bit and 5bit GGML models for CPU inference in llama.cpp](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-GGML)
31
- * [float16 unquantised model for GPU inference and further conversions](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-HF)
 
 
 
 
 
 
 
 
 
 
 
32
 
33
- ## THE FILES IN MAIN BRANCH REQUIRES LATEST LLAMA.CPP (May 19th 2023 - commit 2d5db48)!
34
 
35
- llama.cpp recently made another breaking change to its quantisation methods - https://github.com/ggerganov/llama.cpp/pull/1508
36
 
37
- I have quantised the GGML files in this repo with the latest version. Therefore you will require llama.cpp compiled on May 19th or later (commit `2d5db48` or later) to use them.
38
 
39
- For files compatible with the previous version of llama.cpp, please see branch `previous_llama_ggmlv2`.
 
 
 
 
 
 
 
 
 
40
 
41
  ## Provided files
42
- | Name | Quant method | Bits | Size | RAM required | Use case |
43
  | ---- | ---- | ---- | ---- | ---- | ----- |
44
- `gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_0.bin` | q4_0 | 4bit | 40.8GB | 43GB | 4-bit. |
45
- `gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_1.bin` | q4_0 | 4bit | 44.9GB | 47GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
46
- `gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_0.bin` | q5_0 | 5bit | 44.9GB | 47GB | 5-bit. Higher accuracy, higher resource usage and slower inference. |
47
- `gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_1.bin` | q5_1 | 5bit | 49.0GB | 51GB | 5-bit. Even higher accuracy, higher resource usage and slower inference. |
 
 
 
 
 
 
 
 
 
 
 
48
 
49
- Note: no q8_0 will be provided as HF won't allow uploading of files larger than 50GB :)
50
 
51
- I am investigating other methods, eg a split ZIP file, and will try to upload this soon.
 
 
 
 
 
 
 
 
 
 
 
 
 
52
 
53
  <!-- footer start -->
54
  ## Discord
@@ -70,19 +118,24 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
70
  * Patreon: https://patreon.com/TheBlokeAI
71
  * Ko-Fi: https://ko-fi.com/TheBlokeAI
72
 
73
- **Patreon special mentions**: Aemon Algiz, Dmitriy Samsonov, Nathan LeClaire, Trenton Dambrowitz, Mano Prime, David Flickinger, vamX, Nikolai Manek, senxiiz, Khalefa Al-Ahmad, Illia Dulskyi, Jonathan Leane, Talal Aujan, V. Lukas, Joseph William Delisle, Pyrater, Oscar Rangel, Lone Striker, Luke Pendergrass, Eugene Pentland, Sebastain Graf, Johann-Peter Hartman.
 
 
74
 
75
  Thank you to all my generous patrons and donaters!
 
76
  <!-- footer end -->
77
- # Original model card
78
 
79
- This repo provides the training checkpoint of LLaMA on the alpaca_data_gpt4 dataset via LoRA [MLP] on 8xA100(80G).
 
 
 
80
 
81
  He et al. 2022 gave an insight that FFN can better utilize modification at larger capacities.
82
 
83
  The codes is provided by [tloen/alpaca-lora: Instruct-tune LLaMA on consumer hardware (github.com)](https://github.com/tloen/alpaca-lora).
84
 
85
- We modify the running scripts to
86
  ```bash
87
  torchrun --nproc_per_node=8 finetune.py \
88
  --base_model '/cache1/chtan/large_models/llama-hf/llama-65b' \
@@ -106,7 +159,7 @@ torchrun --nproc_per_node=8 finetune.py \
106
 
107
  **Instruction**: Tell me about alpacas.
108
 
109
- **gpt4-alpaca-lora_mlp-65b**:
110
 
111
  Alpacas are small, domesticated mammals that are closely related to llamas. They are native to the Andes Mountains of South America, primarily in Peru, Bolivia, and Chile. These animals have been domesticated for thousands of years and were used by the Incas for their fleece, meat, and as pack animals.
112
 
 
1
  ---
2
+ inference: false
3
  license: other
 
 
 
4
  ---
5
+
6
  <!-- header start -->
7
  <div style="width: 100%;">
8
  <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
 
17
  </div>
18
  <!-- header end -->
19
 
20
+ # Chtan's GPT4 Alpaca LoRA MLP 65B GGML
21
 
22
+ These files are GGML format model files for [Chtan's GPT4 Alpaca LoRA MLP 65B](https://huggingface.co/chtan/gpt4-alpaca-lora_mlp-65b).
23
 
24
+ GGML files are for CPU + GPU inference using [llama.cpp](https://github.com/ggerganov/llama.cpp) and libraries and UIs which support this format, such as:
25
+ * [text-generation-webui](https://github.com/oobabooga/text-generation-webui)
26
+ * [KoboldCpp](https://github.com/LostRuins/koboldcpp)
27
+ * [ParisNeo/GPT4All-UI](https://github.com/ParisNeo/gpt4all-ui)
28
+ * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python)
29
+ * [ctransformers](https://github.com/marella/ctransformers)
30
 
31
  ## Repositories available
32
 
33
+ * [4-bit GPTQ models for GPU inference](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-GPTQ)
34
+ * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-GGML)
35
+ * [Unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/TheBloke/gpt4-alpaca-lora_mlp-65B-HF)
36
+
37
+ <!-- compatibility_ggml start -->
38
+ ## Compatibility
39
+
40
+ ### Original llama.cpp quant methods: `q4_0, q4_1, q5_0, q5_1, q8_0`
41
+
42
+ I have quantized these 'original' quantisation methods using an older version of llama.cpp so that they remain compatible with llama.cpp as of May 19th, commit `2d5db48`.
43
+
44
+ They should be compatible with all current UIs and libraries that use llama.cpp, such as those listed at the top of this README.
45
+
46
+ ### New k-quant methods: `q2_K, q3_K_S, q3_K_M, q3_K_L, q4_K_S, q4_K_M, q5_K_S, q6_K`
47
 
48
+ These new quantisation methods are only compatible with llama.cpp as of June 6th, commit `2d43387`.
49
 
50
+ They will NOT be compatible with koboldcpp, text-generation-ui, and other UIs and libraries yet. Support is expected to come over the next few days.
51
 
52
+ ## Explanation of the new k-quant methods
53
 
54
+ The new methods available are:
55
+ * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
56
+ * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
57
+ * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
58
+ * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
59
+ * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
60
+ * GGML_TYPE_Q8_K - "type-0" 8-bit quantization. Only used for quantizing intermediate results. The difference to the existing Q8_0 is that the block size is 256. All 2-6 bit dot products are implemented for this quantization type.
61
+
62
+ Refer to the Provided Files table below to see what files use which methods, and how.
63
+ <!-- compatibility_ggml end -->
64
 
65
  ## Provided files
66
+ | Name | Quant method | Bits | Size | Max RAM required | Use case |
67
  | ---- | ---- | ---- | ---- | ---- | ----- |
68
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q2_K.bin | q2_K | 2 | 27.33 GB | 29.83 GB | New k-quant method. Uses GGML_TYPE_Q4_K for the attention.vw and feed_forward.w2 tensors, GGML_TYPE_Q2_K for the other tensors. |
69
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q3_K_L.bin | q3_K_L | 3 | 34.55 GB | 37.05 GB | New k-quant method. Uses GGML_TYPE_Q5_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K |
70
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q3_K_M.bin | q3_K_M | 3 | 31.40 GB | 33.90 GB | New k-quant method. Uses GGML_TYPE_Q4_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K |
71
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q3_K_S.bin | q3_K_S | 3 | 28.06 GB | 30.56 GB | New k-quant method. Uses GGML_TYPE_Q3_K for all tensors |
72
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_0.bin | q4_0 | 4 | 36.73 GB | 39.23 GB | Original llama.cpp quant method, 4-bit. |
73
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_1.bin | q4_1 | 4 | 40.81 GB | 43.31 GB | Original llama.cpp quant method, 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
74
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_K_M.bin | q4_K_M | 4 | 39.28 GB | 41.78 GB | New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q4_K |
75
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q4_K_S.bin | q4_K_S | 4 | 36.73 GB | 39.23 GB | New k-quant method. Uses GGML_TYPE_Q4_K for all tensors |
76
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_0.bin | q5_0 | 5 | 44.89 GB | 47.39 GB | Original llama.cpp quant method, 5-bit. Higher accuracy, higher resource usage and slower inference. |
77
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_1.bin | q5_1 | 5 | 48.97 GB | 51.47 GB | Original llama.cpp quant method, 5-bit. Even higher accuracy, resource usage and slower inference. |
78
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_K_M.bin | q5_K_M | 5 | 46.20 GB | 48.70 GB | New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q5_K |
79
+ | gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_K_S.bin | q5_K_S | 5 | 44.89 GB | 47.39 GB | New k-quant method. Uses GGML_TYPE_Q5_K for all tensors |
80
+
81
+
82
+ **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
83
 
84
+ ## How to run in `llama.cpp`
85
 
86
+ I use the following command line; adjust for your tastes and needs:
87
+
88
+ ```
89
+ ./main -t 10 -ngl 32 -m gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_0.bin --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "### Instruction: Write a story about llamas\n### Response:"
90
+ ```
91
+ Change `-t 10` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`.
92
+
93
+ Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
94
+
95
+ If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
96
+
97
+ ## How to run in `text-generation-webui`
98
+
99
+ Further instructions here: [text-generation-webui/docs/llama.cpp-models.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/llama.cpp-models.md).
100
 
101
  <!-- footer start -->
102
  ## Discord
 
118
  * Patreon: https://patreon.com/TheBlokeAI
119
  * Ko-Fi: https://ko-fi.com/TheBlokeAI
120
 
121
+ **Special thanks to**: Luke from CarbonQuill, Aemon Algiz, Dmitriy Samsonov.
122
+
123
+ **Patreon special mentions**: Oscar Rangel, Eugene Pentland, Talal Aujan, Cory Kujawski, Luke, Asp the Wyvern, Ai Maven, Pyrater, Alps Aficionado, senxiiz, Willem Michiel, Junyu Yang, trip7s trip, Sebastain Graf, Joseph William Delisle, Lone Striker, Jonathan Leane, Johann-Peter Hartmann, David Flickinger, Spiking Neurons AB, Kevin Schuppel, Mano Prime, Dmitriy Samsonov, Sean Connelly, Nathan LeClaire, Alain Rossmann, Fen Risland, Derek Yates, Luke Pendergrass, Nikolai Manek, Khalefa Al-Ahmad, Artur Olbinski, John Detwiler, Ajan Kanaga, Imad Khwaja, Trenton Dambrowitz, Kalila, vamX, webtim, Illia Dulskyi.
124
 
125
  Thank you to all my generous patrons and donaters!
126
+
127
  <!-- footer end -->
 
128
 
129
+ # Original model card: Chtan's GPT4 Alpaca LoRA MLP 65B
130
+
131
+
132
+ This repo provides the training checkpoint of LLaMA on the alpaca_data_gpt4 dataset via LoRA [MLP] on 8xA100(80G).
133
 
134
  He et al. 2022 gave an insight that FFN can better utilize modification at larger capacities.
135
 
136
  The codes is provided by [tloen/alpaca-lora: Instruct-tune LLaMA on consumer hardware (github.com)](https://github.com/tloen/alpaca-lora).
137
 
138
+ We modify the running scripts to
139
  ```bash
140
  torchrun --nproc_per_node=8 finetune.py \
141
  --base_model '/cache1/chtan/large_models/llama-hf/llama-65b' \
 
159
 
160
  **Instruction**: Tell me about alpacas.
161
 
162
+ **gpt4-alpaca-lora_mlp-65b**:
163
 
164
  Alpacas are small, domesticated mammals that are closely related to llamas. They are native to the Andes Mountains of South America, primarily in Peru, Bolivia, and Chile. These animals have been domesticated for thousands of years and were used by the Incas for their fleece, meat, and as pack animals.
165